Overview

CorGeneRF is a platform dedicated to multiomics relationships. It aims to provide users with multiomics features related to clinical factor, the distribution of multiomics features, imaging features related to multiomics features, and multiomics features related to imaging features.

Method

The RNA-seq gene expression with transcripts per million (TPM) are obtained from the illumina sequencing of the colorectal cancer (CRC) tissue. The protein data are obtained by DIA method. The metabolisms data are obtained by DDA method. The imaging data was derived from the CT DICOM data segmented by the 3D slicer. The gene expression of normal and tumor tissue was exhibited with histogram. The pathway scores were derived from the GSVA analysis. The correlation of genes, proteins, metabolisms, pathways and imaging features were assessed with the Spearman correlation. The expression level of genes were classified by the model developed with significantly correlated imaging features with logistic regression (LR) and least absolute shrinkage and selection operator (LASSO). The significantly different molecules in different clinical phenotypes were calculated by “DESeq2” package for RNA, by “limma” package for protein, by wilcox.test for metabolism, by foldchange with the ratio of average of different groups for imaging feature. And the relationship of molecules with clinical factor were showed in the KM curve.

Guideline

    Our website enables you to explore the correlations among genes, proteins, metabolites, pathways, and imaging features. The information that can be queried includes clinical characteristics (overall survival, metastasis, stage, gender, age, colorectal cancer subtype), multiomics features (9965 RNAs, 5449 proteins, 637 metabolites), and imaging features. The features that can be queried on this website are listed in the Items table on the dataset page. Features not in this table cannot be queried.

    You can also investigate the expression levels of genes, proteins, and metabolites in tumor and control tissues, identify the distinct molecules associated with different clinical factors, conduct Kaplan - Meier (KM) analysis on the top - ranked differentially - expressed molecules based on high and low levels, and view the ROC curve for predicting gene expression levels using imaging features.

    On the ‘Home’ page, you can select a gene symbol, protein, metabolite, clinical factor, or imaging feature, then enter the corresponding name into the search box. Subsequently, the analysis results will be presented on the ‘Search’ page. The download documents of the analysis results, including the pdf file and the table of the analysis results, will be displayed in the lower-left corner of the ‘Search’ page. Below, we will use clinical factor (OS), molecule (CD4), imaging feature (original_shape_Flatness) as examples to demonstrate how to navigate our website.